|||
Note: These are the prepared remarks I wrote for my talk at IxDA Oslo. In the end, I spoke more extemporaneously, but I thought at least I could share my notes to give you something closer to what I was getting at in my q&a remarks. Ultimately a meetup was a great place to debut a new idea, I’m looking forward to writing — and talking — more about these ideas over the next year.
Systems fail. They falter, decay, break under pressure, or never quite work in the first place. But most of the time, they do not fail all at once. They leak, jam, get stuck. And when that happens, it is not designers or executives who get the call. It is the people who live inside those systems: They are the ones who keep things running.
Let me give you examples of what I mean.
These are acts of care. They are also systems interventions. And they expose a reality that most design frameworks do not want to admit: The most important part of a system is often the person who keeps it from breaking today.
This work has always existed. It is the daily practice of making broken systems barely functional. It is the mechanics of how public life actually operates. The people and practices that keep things running despite institutional indifference.
What is needed now is not more big ideas or breakthrough products. It is repair. Design as repair means paying attention to what is already in motion. The fragile, fragmented, overburdened systems we inherit. And asking: How do we maintain this? Who holds it together? What makes it fail?
Mark Fisher argued that we use technology to endlessly simulate the twentieth century rather than imagine new futures. I would argue it is even more specific: we use technology to patch the crumbling infrastructures of the twentieth century while pretending it is progress.
Design as repair is design without the hero narrative. It is design that starts from consequence, not control. From entanglement, not abstraction.
And this matters because the dominant posture in tech and design is still one of optimization and solutioning. A logic that erases the embedded work of upkeep and misreads care as inefficiency.
Ironically, AI has made this maintenance work more visible. We are now flooded with tools that promise automation and end up producing more repair work.
Writers are hired to rewrite AI-generated content that lacks clarity or context. Designers are paid to fix broken AI logos or redraw pixel soup into something usable. Engineers are tasked with cleaning up buggy AI-generated apps.
The loop is clear. Automation without understanding leads to more friction, not less. And in every one of these cases, human judgment and care becomes the final defense.
It is way easier to leave after the site has been shipped, because that is when the real work begins. But someone always stays behind to deal with the mess. And it is rarely the people who created it.
We used to study systems seriously. Cybernetics, operations research, institutional analysis. Now we have systems thinking workshops and feel-good frameworks. But systems do not care about our mental models. They work or they do not. People eat or they do not. Services reach people or they do not.
Erika Hall has pointed out that the real problem with Design Thinking wasn’t the sticky notes or workshops, but that it became a vending machine for “human-centered innovation” that ignored structural conditions. It confused a consulting toolkit for a curriculum and left designers unprepared for messy, real-world consequences.
Anis Punjwani describes a parallel trajectory from the Global South: design education promised rigor and social impact, but the industry reduced it to tool use and aesthetic homogenization. Research took a back seat, and critical practice was subsumed by tech’s obsession with speed, novelty, and now AI.
What we need is practical knowledge. Understanding how things actually function through the accumulated wisdom of people who show up every day and make broken things work.
James C. Scott wrote about how states try to make societies legible. Turning complex, local realities into simple categories they can manage from a distance. The problem is that this legibility is always a reduction. It misses the knowledge that actually makes systems work.
This is not a call for more empathy. Empathy has limits, especially when systems are indifferent to the feedback we collect.
What we need is infrastructure. Not just pipes and cables, but institutional slack. The space for people to do the right thing. The permission to stop the line, make adjustments, ask questions, or delay the next step. That is not inefficiency. That is how things do not break.
But here is what we often miss. Every one of these interventions requires coordination. Someone has to notice the problem, escalate it, get permission to act, find the resources to fix it. Someone has to communicate the change, update the documentation, train the next person. Andrea Magnorsky has written extensively about how coordination is not free, and decision-making reflects the power structures and feedback loops embedded in our systems. If we care about repair, we must also care about how decisions are made, who has context, and who is allowed to act on it. This is not a UX problem. It is a governance problem disguised as design.
The problem with much of today’s design talk, especially in government and enterprise, is that it is still overly fixated on the interface layer. We make better dashboards while the underlying eligibility logic or staffing policy remains untouched.
We need to stop treating design as the surface and instead embrace what Hamid Ekbia calls the distributed burden of heteromated systems. The question is not whether a system is human-centered, but whether it creates the structural space for care, context, and correction.
What happens when we treat the system itself as the object of design? When we design for the caseworker’s workaround, not against it? When we account for the bus driver’s local knowledge instead of optimizing it away?
For those of us who call ourselves designers, this is both humbling and clarifying. We are not the heroes of this story. We are not even the most important actors. But we can be useful.
We can make the invisible work of care visible. We can document the workarounds that keep systems running. We can advocate for the resources and authority that care workers need. We can design processes that create space for human judgment.
The real challenge is not designing better interfaces. It is designing organizations, policies, and processes that can maintain care over time, even when they are under pressure.
Instead of the forward-deployed consultant who parachutes in with solutions, we need designers embedded in the operational reality where systems actually break and get repaired. People who stay with systems after they ship, who understand how they degrade, who learn from the mechanics keeping them alive.
We need new models of professionalism that do not center novelty, but consequence. New ways of valuing labor that include the cost of wasted time, operational debt, and institutional indifference. New feedback cultures that reward slack, not just efficiency. And new design approaches that understand power not just as influence, but as structural access. The ability to change the roadmap, to override a rule, to slow down.
The mechanics are already out there, keeping things running. Our job is to see them, support them, and design systems worthy of their care.
Because if we are serious about building a better world, it will not be through clean handoffs or pixel-perfect flows. It will be through janitors, drivers, support agents, and system stewards.
Design is repair. Let us act like it.
Thanks to Fredrik Matheson and the entire IxDA Oslo team for their hospitality and professionalism, it’s constantly one of the best run meetups in the world.